Originally Posted by Blameless
I've done platter swaps on somewhat older drives for the purpose of getting data back that was worth more than a drive but less than the thousands it would take to send it to professionals with a proper clean room and the right tools.
Sometimes the drives work long enough to recover something, some times they don't, and they almost never work again, but if you understand this (and have already tried a PCB swap), it can be vaguely viable.
I do think the need for alignment of the heads/platters is overstated. Small deviations should be corrected automatically, else normal thermal expansion and contraction would cause drives to become non-functional, instead of just needing a recalibration retry.
Of course, I haven't attempted anything like this with modern PMR drives, or those with multiple platters.
The multiple platters part is what would kill it. I could actually somewhat see it being -sort of viable- on a single platter drive, if you just wanted to try the theory out on a dead drive.
But IIRC, the platter alignment spans across all of the platters, so you'd have to put them on their new spindle in the exact way that they were removed from the original. Else the drive would probably end up mixing clusters up. Terrible at trying to express what I mean, but think RAID 0, with fixed placement of data. Where you took the drives out of the array, and then put them back in a different order. The controller would get blocks all mixed up, because it would look for block A on drive A, and then say, block B for that file would end up being on drive C, but the controller would expect drive B. (Yes, I know RAID doesn't quite work like this).